Master M2: Representing 4D human motion sequences using dance notation

We are looking for a highly motivated Master student of a M2 internship.

Context:

Dance notation systems provide a representation of human motion, mostly used in modern dance, that is localized in space and time [1]. These representations have recently been used in the fields of robotics [2], computer graphics [3] and cultural heritage [4] to represent human motion. In this internship, we are interested in exploring the suitability of dance notation systems to represent human motion captured at high frame-rates and and high spatial resolution in a multi-camera system. We will explore whether the representation allows for robust extraction and faithful yet compact representation of general motion sequences. In practice, we work with the Kinovis acquisition platform located at Inria Grenoble [5], which uses 68 synchronized color cameras for the acquisitions.

Left: example sequences of human motion color-coded by the time at which frames appear in the sequence. Right: Corresponding dance notation.

Environment:

The work will take place within Morpheo team at Inria Grenoble Rhône-Alpes. Inria is a leading French research centre in computer science.

Objectives:

  • Study the relevant bibliography and dance notation implementations
  • Implement a solution to represent data captured from a multi-view studio using dance notation
  • Analyse the results and validate the representation using (a) motions with known notation, (b) different characters performing same type of motion (e.g. walking or same dance steps)
  • Use the representation to animate a new character with the same motion
  • Possibility of capturing test data using the Kinovis platform
  • Write a Master’s thesis with details of the proposed method, with bibliography and experiments

Student profile:

  • Master student – preferably in Computer Science or Applied Mathematics.
  • Creative and highly motivated
  • Solid programming skills, Python, C++ and/or Matlab
  • Solid mathematics knowledge in linear algebra, geometry, and statistics
  • Fluent English or French spoken
  • Prior courses or knowledge in the areas of computer vision, computational geometry, mesh processing, computer graphics, signal processing, machine learning or Bayesian inference is a plus

Duration: 5-7 months

Start date: February 2022.

How to apply:

The internship is co-supervised by Stefanie Wuhrer, Joao Regateiro and Sergi Pujades. Informal inquires can be addressed to stefanie.wuhrer@inria.fr. Please apply on the INRIA website (we will make the link available soon).

References:

  1. Labanotation The system of Analyzing and Recording Movement, Fourth Edition. Ann Hutchinson Guest. Routledge 2005.
  2. Describing Upper-Body Motions Based on Labanotation for Learning-from-Observation Robots. Ikeuchi K., Ma Z., Yan Z. et al.  Int J Comput Vis 126, 1415–1429 (2018).
  3. PERFORM: Perceptual Approach for Adding OCEAN Personality to Human Motion Using Laban Movement Analysis. Durupinar F., Kapadia M., Deutsch S., et al., ACM Transactions on Graphics, Vol. 36, No. 1, Article 6, (2016).
  4. Labanotation Generation From Motion Capture Data for Protection of Folk Dance. J. Wang, Z. Miao, N. Xie, W. Xu and A. Li. IEEE Access, vol. 8, pp. 154186-154197 (2020).
  5. https://kinovis.inria.fr/

Comments are closed.